Hey, I'm
Geetheswar
Data Science and Engineering Student at IIT Palakkad.
I build scalable AI-powered products — from modern web applications to autonomous backend systems — blending machine learning, automation, and efficient cloud engineering.
Designing intelligent systems with a builder's mindset.
Framing the journey from data science student to production automation.
I'm a final-year B.Tech student in Data Science and Engineering at the Indian Institute of Technology, Palakkad. I have a 6-month internship experience at Brix AI Technologies.
During my internship, I've contributed across the full stack: crafting frontend interfaces, developing robust backend APIs, managing CI/CD pipelines, and maintaining production deployments on VPS and Docker environments. I've also designed automation flows using N8N, secured application endpoints, and optimized internal DevOps workflows.
My interests span AI agents, LLM-based applications, cloud-native systems, and data-driven automation — especially where they combine to create reliable, scalable, real-world products.
A dependable toolkit for intelligent products.
Combining software, infrastructure, and machine intelligence to ship scalable systems.
Languages
Frontend
Backend & Services
AI & Data
DevOps & Infra
Messaging & Caching
Datastores
Automation
Practicing hands on AI-powered development and scalable systems.
My experience through internships and projects in building AI-powered applications and scalable backend systems.
Company
Brix AI Technologies
Jun 2025 - Nov 2025
AI & Full-Stack Intern
Contributed to developing AI-driven healthcare automation systems integrating LLM pipelines and scalable backend architectures.
- Built dynamic front-end modules using Next.js and Tailwind.
- Developed REST APIs and backend automation with FastAPI.
- Managed multi-service Docker deployments across VPS environments.
- Automated workflows using N8N for backend data orchestration.
- Implemented CI/CD pipelines for production releases.
- Handled application security reviews and monitoring.
Most Recent Works with production grade engineering.
The products and platforms I've built recently.
SagePick
Led the development of a microservice-driven platform that delivers real-time recommendations for movies, series, and anime by pairing granular watchlist tracking with preference models and automated catalogue updates.
Designed an AI discovery engine that uses agentic search across public databases and long-form articles to surface titles that match scenarios or moods described by the user.
Orchestrated user_service, core_service, and ai_service inside a shared Docker environment with RabbitMQ, Redis, and three databases, enabling event-driven retraining, catalogue refreshes from TMDb, and elastic scaling as engagement changes.
Context Aware Science Assistant
Built a configurable RAG stack that streams academic context from arXiv, Scholar, and more, then reasons with hybrid retrieval and rerankers to answer deep-dive research prompts.
Reticular Material Discovery
Explored generative flow networks to design CO2-adsorbing reticular materials, coupling topology constraints with interactive 3D inspection workflows.
Hokomi Tracker
This project is a personal media tracking platform built with Next.js, React, and Prisma, helping users manage and document their anime and manga experiences. It allows users to track watching and reading progress, record personal notes, and link them to specific streaming or reading platforms. The application integrates the Jikan API to fetch detailed anime and manga data, ensuring accurate and dynamic content updates. Developed as a foundational full-stack project, it demonstrates skills in database modeling, API integration, and responsive frontend design. Future iterations will include enhanced personalization, advanced analytics, and improved UI/UX.
Dynamic Weather Analytics
This project serves as my foundational exploration of Data Analysis, Machine Learning, FastAPI, and ReactJS. It enables users to input weather parameters such as humidity and temperature, automatically generating analytical insights through graphs and text summaries. Users can also train machine learning models to predict specific features using others, showcasing the application’s analytical depth. Built with a FastAPI backend and a React frontend, it demonstrates skills in data visualization, model integration, and interactive analytics, with future plans to enhance automation and expand predictive capabilities.
Let's Connect
Reach out for any quries and collaborations.
Ready to start a project?
Drop me a line and let's discuss your ideas.